Introducing GBD 2021

The Global Burden of Disease Study (GBD) is the most comprehensive assessment of health trends and conditions across countries. The latest GBD studies provide detailed analysis of disease burden related to life expectancy, fertility, neurology, and many other health topics. We discuss GBD with IHME Director Dr. Christopher Murray.

This transcript has been lightly edited for clarity

Rhonda Stewart: Welcome to Global Health Insights, a podcast from IHME, the Institute for Health Metrics and Evaluation. I’m Rhonda Stewart. In this episode, we’ll hear from IHME Director Dr. Christopher Murray as he talks about the Global Burden of Disease study, also known as GBD. GBD is the largest and most detailed scientific effort undertaken to quantify health trends. GBD provides a unique platform to compare the magnitude of diseases, injuries, and risk factors across age groups, sexes, countries, regions, and time. For decision-makers, GBD provides a unique way to compare countries’ health progress and to understand factors that impact health, such as high blood pressure, cancer, heart disease, and many others. Led by IHME at the University of Washington, GBD is a truly global effort, with more than 11,000 researchers from over 160 countries participating in the research. The latest GBD studies are coming out now and cover a wide range of topics, including fertility, life expectancy, risk factors, and others.  

Chris, the Global Burden of Disease study has a long history, but it’s worth talking about why the study was created in the first place and what GBD aims to do in terms of its comprehensive nature. So can you talk a little bit about that?  

Chris Murray: Yeah, Rhonda, the GBD started back in 1991, and it was born of the interest of the World Bank, actually, in having a more comprehensive view of what the health problems were in the world for their first-ever bank-wide policy document on health called the World Development Report in 1993.  

The work started a couple of years before that report was to be released. At the time, I started work on how we would do that, how would we get a comprehensive view. And if you went around the World Health Organization in 1991 and added up the number of deaths different programs claimed, those deaths were about three times larger than the total number of deaths in the world. So there was a real mismatch. And it was because basically data was being driven by advocacy needs, and programs at the time wanted to make their diseases or even their risk factors look as large as possible because they thought that was a strategy to get more funding.  

That advocacy-driven environment meant that there wasn’t somebody who just wanted to know the facts, you know, what was the biggest cause of death? What was the biggest cause of ill health? There wasn’t anywhere to go to, and thus was born the idea of a comprehensive view of that, that reflected the data that was out there, used standardized methods and required that people only die once, that were dying, and then went a step further and said, we don’t want to just know about death or premature death, we want to understand illness, loss of health function, so that diseases that didn’t necessarily kill a lot of people but were important in people’s lives would also be captured.  

And that commitment to a comprehensive view, a comparable view, one that had what at the time was thought about as parity between disabling conditions and fatal conditions were all part of the reason the GBD was created and the reason that it caught on and has been running for 33 years.

RS: And so what makes GBD relevant today still, after 33 years?  

CJLM: Well, the exact same reasons that the GBD was created in the first place, namely, the need for scientifically rigorous, comparable information that gives you the big picture on health problems. There isn’t anywhere else to go to if you want to understand what are the big contributors to ill health in a place, how are they changing over time? How does my community stack up compared to other communities?  

And, you know, it’s back to the old adage that you manage what you measure. So unless you have measures comprehensively of health, it’s very easy for governments and other actors to end up ignoring a problem – let’s say mental health. Before the GBD came along, lots and lots of countries didn’t even have a policy about mental health because there was no information on how big a problem it was in their community. And the same has turned out to be true for other conditions. So that that sort of grounding in what does the data tell us really changes the nature of discourse, both at the local level, the national level, and certainly at the global level.  

RS: IHME coordinates the GBD, but the work is carried out by a global network of researchers. What does their participation add to the study, especially in terms of subnational studies around the world and in the US?  

CJLM: Well, particularly in the last 15 years, we have invested a tremendous amount of effort in building a very inclusive group of co-investigators for the GBD. We now have more than 11,000 collaborators who have read the study protocol, agreed to its terms, and signed on to contribute, and that 11,000-plus individuals from 160-plus countries really strengthens what we do. It helps us identify data sources that we might have missed just looking globally or in the literature. They provide sort of ground truth, common sense, you know, local knowledge about which sources should be used and what are some of the biases in different types of data sources. They help in the statistical model fitting and interpretation, and they write policy documents and research papers. They become local champions for the analysis to drive the use of the GBD in low-, middle-, and high-income countries.  

One of the best manifestations of that spirit of collaboration and the network we built is the countries where we have subnational burden of disease work. So in a country like Brazil, there’s a very active network of collaborators. They push the GBD down to the state level in Brazil. They’ve used that state-level burden in national and state-level planning. They’ve written 100-plus papers on burden in Brazil. And it’s an example where even with changes in government, we’ve got such a good, active group of collaborators that that work has just kept going and is being used by governments of different persuasions to help inform policy formulation.

The US is another example where with funding from NIH [National Institutes of Health], we’ve been able to take burden of disease not only to the state level, but now going to the county level and now county and race-ethnicity groups. So we’re analyzing the burden of disease with NIH down to more than 15,000 units of analysis in the US. So the network of collaborators, the way it’s grown in certain countries to be so much more detailed, also enhances its relevance and utility for decision-making, priority-setting, understanding disparities, lots and lots of ways in which more grounded, empirical, comparable information can change the nature of discussion in the country about what are our priorities.

RS: Let’s talk about the latest version of GBD and explain some of the factors that contributed to the timing of the latest release of the data.  

CJLM: So it’s 2024, and we are putting out results in the GBD from 1990 through to 2021. And that’s a more of a delay in the release of our results than historically has been the case. And it’s because of the complexity of COVID. The effects of COVID have been really different in different countries, even countries that are neighbors. They’ve been very different across age groups, and they’ve played out in different countries as a function of both the individual behavioral response to the pandemic – you know, government policies to limit transmission, what some people call non-pharmaceutical interventions – as well as the scale-up and rollout of vaccination and in fact, the access or lack of access in the case of many low-income countries to vaccination.  

All of those have materially changed how COVID has impacted different places. And it’s taken us quite a long time to get enough data in because of lags in data systems to feel confident about understanding what the first two years of the pandemic actually looked like. I’ll give you some examples of that. We’ve seen really large effects in the deaths per 100,000, 400,000, how you want to measure them in places like Bolivia, Peru, Ecuador, some of the states in Mexico, some of the parts of the US, for example, the Navajo Nation, where the effects of COVID just in the death count have been really large.

And then we’ve seen, with less tremendous data, but still enough data that we were pretty convinced, that once you control for age, the effects of the pandemic have been really quite large in sub-Saharan Africa. It’s not immediately obvious because the age structure is so young, but we think – and South Africa certainly shows it to be the case – that there have been large effects in older age groups, even though they’re a small percentage of the population in sub-Saharan Africa. At the same time that we’re seeing these effects, it seems like child mortality kept going down. And so, in fact, child mortality from flu, from some of the other respiratory pathogens, measles, diphtheria was reduced because of the behavioral actions, the non-pharmaceutical interventions that reduced those pathogens and were implemented because of COVID.

So super complicated mixed effects. Getting the age structure right has been a challenge. And then you have places that kept their borders closed for most of 2019, 2020, and 2021, like Australia or New Zealand or Singapore. That really meant that actually mortality was reduced during those years rather than increased because they didn’t have the COVID deaths very much at all in those years. They did in 2022, but they had the reduced effects of, for example, flu in adults, because of the reductions in pathogen transmission. So much more complicated, heterogeneous effects, age-specific effects, very different than we typically see from year to year in the GBD. And that’s been part of the challenge of getting this study done in a rigorous way.

RS: And so you just talked about some of the things that the latest study tells us about COVID that are new, things that were not known before. And can we expect that GBD will continue to provide information about COVID that hasn’t yet been released or hasn’t yet been known? Is there still more to be told when it comes to the story of COVID?

CJLM: There is a lot more to be told on the COVID story or the pandemic story. And as we’re now fully into the cycle of peer review – you know, there’s a lag from when we finish the analysis to when it gets published because we go through a very rigorous peer review process and collaborator review. So we’re already well into the next cycle of GBD, and we’ve been looking at the data as it’s coming in for 2022 and 2023, and that COVID story is still going to be surprising to people.

We’re seeing countries that had very low COVID mortality in 2020 and 2021 now faring rather poorly, and we’re seeing the countries that people pointed to as success stories – Northern Europe in many cases is an example, Canada, Australia, New Zealand, [South] Korea – now having a much harder time in more recent years. So I think we will continue to learn about the complexity of how the pandemic has played out and the myriad pathways by which it’s affected health and mortality: deferred care, rising drug use, rise of depression and anxiety in certain age groups in certain places.

Quite an incredible social shock from the pandemic, and it’s now going to take many years, I think, before we fully understand all the nuance of how the COVID pandemic has wreaked havoc on health and society around the world.  

RS: And, Chris, there are numerous examples of how GBD has been used by policymakers and others. What value does the study have for evidence-based decision-making? You talk about this a little bit in describing the subnational studies and also the roots of GBD. Tell us a little bit more about this idea of how this information can really be used for action.  

CJLM: So the most powerful thing that the GBD provides, and we see it in country after country in their strategic documents where they use the GBD on the legislation that gets pushed through by governments, is agenda setting, where having a good comparable notion of what are the big problems in a society and which ones are getting worse really helps governments set agendas.

And we see this in low-income countries all the way through to high-income countries – you know, Public Health England, when it was still an independent agency, as an example, used the GBD to set their strategic priorities. We see the same thing happening in middle-income countries, in low-income countries, having a notion of what’s a big problem, is it getting better or worse, turns out to be enormously useful for setting what should be toward the top of the agenda.

Specific analyses have driven governments to take specific action, so the finding in the GBD of how many kids are dying from indoor air pollution has led some countries to adopt policies around cleaner indoor fuels like clean cookstoves. In other countries, the GBD is showing that there are a lot of deaths related to alcohol, and their going up has led to taxation legislation to raise taxes on alcohol.

As we as we talk to governments around the world, we find out that they have found many, many ways to use the GBD to help set priorities. And as we’ve added in the latest round, and is in the last of the six papers for GBD 2021, forecasting out to 2050 with both what we think is likely to happen and alternative scenarios. Those also turn out to be very useful. Like how big of an effect will climate have in my country? What does it look like going forward? What are some of the opportunities if I address certain major risk factors to substantially change that trajectory? And that’s now built into the forecasting part of the GBD as well. So it’s not just an assessment of what’s been happening in the last 30 years, but it’s also where we might be going in the next 26 years to 2050.

And in fact, in other analyses coming later, we will be seeing the forecasts run out to 2100 because that’s relevant to things like climate. So lots and lots of specific applications of the GBD and surprising innovation from our 11,000 collaborators in how you can use a curated dataset or comprehensive dataset on health for medical education, for health education, for so many different applications as we look into the future.

RS: Great. Thanks so much, Chris.

CJLM: Okay. Thank you for your interest.

RS: Details about the Global Burden of Disease study and a wide range of GBD-related resources can be found at healthdata.org.